Its an interesting topic Adam. There is another solution depending on your needs and the degree of shading that you suffer. My system has 2 banks of 2 panels (2 panels in series to lower transmission losses). The 2 banks are wired in parallel (with blocking diodes). Ideally I would have had a separate controller for each bank - indeed, I've seen an installation with a charge controller per panel and the current trend is towards microinverters. I've had my installation for 10 years now and now that the price point of charge controllers has reduced (along with _some_ having appropriate controls and profiles for LiFePo4) it may be time to reconfigure. That said, I'm still pretty happy seeing 40+ amps charging the batteries after dealing with the fridge, freezer and other loads.
I had a robust argument with a few members of a forum I belong to re. panels in parallel. I advocated fitting a Schottky diode in series with each panel to prevent reverse feeding in shade. The others said don't bother. I conceded defeat and left the diodes out, after a lot of head-scratching. My logical conclusion goes like this: 1. None of my panels will ever be covered by a totally opaque sheet of anything. 2. The graph on the manufacturers data sheet shows that the voltage output of the panel is almost constant from max. insolation to min. insolation, only the current varies. 3. In almost any light, with no load, the voltage is almost the same, whether unshaded or translucently occluded. 4, Once the panels are illuminated, the MPPT regulates the terminal voltage to what it finds best, always below Voc 5, Therefore, if one panel is fully lit, and the others are cloud/tree shaded, the voltage will stay the same across each panel, but the current each panel contributes to the total will be in proportion the the illumination. 6. With one of my panels behind a tree shadow, and the rest fully illuminated, the Voc of the shadowed panel is higher than the Vmp of the other 5. 7. Connecting the shadowed panel in parallel with the others it's voltage drops to match, and marginally increases the total current by a few mA. My experiments proved my original argument on the forum wrong, diodes unnecessary unless the shadow is a cardboard sheet on the surface of the panel. I ate humble pie; the money saved was worth a pint, and 25W loss on a cool bright day. Just my tuppence worth!
Solar panels are just a bunch of crappy diodes in series. Much better than they used to be. Conduction should not happen until voltage gets significantly above the open circuit voltage. In fact, if you have snow on your panels that can be melted off by back feeding voltage as long as the current is less than the panels rated voltage.
Most solar panels have a 1000V diode 1007 kind which makes 0.7V drop or more. Obviously a 40V Schottky Diode can do better given not too many are connected in series. I think the idea of using a 1000V one is also to make this more robust against voltage spikes as they might occur in lightning strikes (which won’t hit your house but something else close enough to have an impact). The mosfets chosen are probably also low voltage kinds with similar risks in case of a spike or so. In case of dozens of solar panels on the roof who wants to periodically check diodes and diagnose might be difficult. The waterproof box probably isn’t waterproof after opening it up twice...
Guess what, Schottky diodes leak too. Especially when hot. Lower voltage rated ones have lower forward voltage than ones with a higher PIV. I was hit by lightning and it took out a number of bypass diodes protecting the panel. Thank heavens for cheap diodes. They were easily replaced and the panels are still working.
Hey, great video again. To my understanding panels themselves are bypass-protected so that a shaded panel does not have to pass the current through it's cells during shading of a panel in a series string. This mimics removing a panel from the series and avoiding the shaded panels resistive properties. However, when "overloading" a solar chargers input with more than one string in parallel, at most with one each facing east, south and west, there'll be a situation where entire strings of panels are shaded. Especially when the voltage of the strings, or difference of voltages of single parallel panels are great enough, damage may occur, to my understanding again. I'm about to combine a 21Voc and a 41Voc panel in parallel on a single MPPT and will be using blocking diodes. To demonstrate, try connecting a 230V lightbulb to a 12Vdc source and observe current; nothing much happens... Until the voltage becomes high enough to overcome the resistance and the wire starts to glow. I'd avoid making panels into lightbulbs... Panels do not contain blocking rectifying on top of bypass, because it would lower the ratings of the panel. They also want to avoid adding cost because they want to be the cheapest on the market... But when you go farther away from the sun to say Finland, you either have to move panels around according to time of day or get more of them in different directions in order to maximize output and usage of an expensive inverter in order to make things as affordable as possible. Maybe there's a sequel to this topic? Thanks and keep up the great work!
Couple of points: most panels seem to be one long series string of 0.5v cells and shading just one cell can seriously reduce output. If one cell is a bit dirty, what happens to output ? Secondly when doing this sort of experiment it would be better to find a constant and thus repeatable load rather than just a battery on charge.
Within panels the same thing happens just on a smaller scale. So yes, one "dirty" cell will reduce output more than just having one cell less. The load is fine in this setup as long as the batteries aren't the limiting factor. MPPT tries to get the most power out of the cells in any case. And comparing measurements more than a few minutes apart is pointless anyway, because the sun/haze will have changed more and uncontrollable than anything else in the setup.
So this is not a video on how to shade panels! Ha-ha great video, alot of ppl think a little shadow on a panel is nothing!! In fact it is a big nothing when it comes to power!!
The parallel connection is very interesting. In this case the consumption of the shaded panel is tiny but when you have strings of 23 modules of 300w each in parallel it may be required to use string fuses to avoid destroying the strings when heavy shading occurs as the Voc will be close to 1kV and the current capabilities are 9-10A... Each manufacturer will give the maximum series fuse rating in amps which is usually 15A or 20A. Usually string fuses are required when more than 2 strings are installed in one MPPT in parallel and they are there to protect the string from being destroyed. Also, string diodes are not used anymore.
So on my pair of panels on my shed roof, one often gets a bit of shade from a neighbour's tree, and later in the day a shadow from next door creeps across them. Looks like I need to parallel them up - even without diodes/ideal diodes, the loss was much less on parallel than serial!
@@ganey I've just rewired my pair to parallel. Bit of a poor day weather wise for solar panels though so will have to wait for the next sunny day to really compare!
It's hard to compare with the randomness of our weather but my pair of panels are definitely working better since changing them to parallel. They produce power later into the day and the peak power is around 10% higher than I've ever achieved before!
Hi Adam and all Thank you for the demonstration of the ideal diode, One question that haven’t been mentioned, is the effect of putting miss matched panels together. And how to avoid the same problem, for example putting a 40volt panel in parallel with a 20volt panel. in this instants the problem you discussed would be much worst, and best avoided by trying to match panels (in parallel for VOLTAGE, and in series for CURRENT) would you agree ? This is something I’m trying to tackle with expanding my solar array and deciding whether or not to add Schottky diode in one of my two strings. Many thanks for the video.
How much power could be gained, by ditching the diodes, for the ideal diodes? On a system with about 10, 300w panels? That's about 5volts, total!! I guess it could help a small system, great video...
hy brother i want to ask about diode ideal..can use it to mppt esmart..i see spesification mppt dont connect solar panel to mppt without battery connected
A PMOS, a 100 Ohm gate resistor, and an optional 18V Zener to protect the gate against over voltage (20V), will do the same job. I think the 6-pin IC that is controlling the 2 PMOSFETs, is a cheap Chinese clone version of something like an LTC4359, that has been repackaged in a 6-pin package. The real ones are 8 pins, but 3 pins have no internal connection. I think the cheapest, with least resistance, is a fuse and diode combo. If the power is reversed, the diode across the power will conduct and blow the main fuse.
Have had this video on my must watch list for a while, glad I did. Very interesting as usual. But, what about low light? Realistically most solar setups are capable of producing "more than enough power in ideal conditions so the differences in setup are not really important until winter when the sun is low and clouded. Would love to see your opinions on low light solar generation variations
dont all panels have internal diodes....And are they not a high resistance so not really a burden to other parallel panels? I need to investigate this as i have not seen this parallel problem before/in my setup. edit ok you just did what i was going to do, and 1/3w loss, meh, i'll live with that
Off topic. Solar related. Eco Worthy now also has a 40 amp MPPT charge controller that seems to have bluetooth communications ability with a SmartPhone. Years ago you reviewed and used their 20 amp MPPT, and I own the 20 Amp version. I have moved on, but I thought their 20 amp MPPT was a great value. I was wondering if their 40 amp MPPT was competitive in the value market.
For a tipical MOSFET (NTB5860NL), the On resistance is 0.003 Ohm so at 1.9 A it will be *6 mV* This is because when the transistor is on it behaves like a resistor, the parameter is Rds(on) if you look on the datasheet.
hey thanks, would you be able to duplicate the test with partial shading? I'm looking to set up a test of single axis versus dual axis small array... Thanks!!!
Thanks Adam. I've only ever made small solar panels from individual cells, and always fitted a suitable Germanium diode (back in the 1970's and 80's) in series with each panel, to avoid reverse-bias situations. Do commercial panels not have similar protection? If not, then it's sensible to fit your own diodes, or MOSFETs, in a parallel array, as you describe here.
I’ve not seen a panel supplied with an inline diode. Usually this is left to the solar charge controller which will have protection on the solar input, but of cause this doesn’t help when paralleling up panels which get shaded. Most have bypass diodes for when they are setup in series though.
Each of my panels is connected to its own microinverter so, in theory, I should benefit from taken the diodes out of the panel altogether, is that right? Also, I'm considering making a solar tracker as I only have 6 250w panels and I believe it is cheaper to make more power from the panels I got rather than buying more panels. Have you considered making a solar tracker yourself?
Those in the panel are bypass diodes. They prevent damaging a cell that is shaded by reverse voltage. If you don't have severe shading they do nothing either way. Remove tham and you will destroy the panel. If you have the room, panels are way cheaper than tracking these days if not buying battery voltage panels.
Thanks for this interesting video. However, I'm confused or maybe missing something obvious. With two panels in series you measure 75W, but when in parallel only get out 72% (2x27W=54W) power in full sunlight (or was the measurement taken in different solar conditions). Why such a difference? Is serial solar panels setup always more power efficient?
That's because it's not a lab test where you can keep the "sun" constant. The position of the sun changed, maybe there was more haze, that already makes a big difference. Comparing measurements a minute apart might be fine, anything more than that not really.
Yeah there was some time between these tests. Conditions were warm and sunny but there were some clouds around. It’s hard to get the exact same conditions for each and every test, but actually it wasn’t really needed for this one - it was more about comparing shaded and unshaded panels.
just opened one of my 38v 300w panels, to find its 2 series 19v 150w panels..... the 'first' panel has a diode ''over it'' to prevent reverse voltage, and the 'second' panel also has this, and the + of the first is connected to a diode which goes to the - of the second... odd but functional way. there only cheap silicon diodes, makes me wonder why decent manufacturers dont change them for something better but i suppose its how its been for 30 years or more
I don't think it's worth the trouble to put ideal diodes on every panel in parallel, and certainly not conventional diodes as they will end up wasting more power than they save. In my setup that would require a *hell* of a lot of ideal diode modules! However I am considering building my own high powered versions to replace the main reverse blocking diodes in the system. One day I'll get a Round Tuit. On the other hand, parallel panel configurations of solar modules are superior if they are in a situation where any part of the array gets shaded during the day. And of course, the only way if you want to have different directional facing modules on the same controller. Five weeks past the Winter Solstice here - the Sun is coming back nicely. :)
Hey Xan. Glad you’ve got through another winter and the sun is returning. I agree, in most cases the diodes are unnecessary and reducing or removing the shading is always going to be the best option. Plus I doubt anyone would experience quite the level of shading I tried here! I tried to cover this in the conclusion at the end of the vid. Cheers for your input as ever.
@@AdamWelchUK Yes, in certain situations diodes are useful. I actually do have some, one of my arrays has a combination of standard 36 cell 12 volt nominal modules and also three bizarre 48 cell modules. These are not grid tie panels, they are actually sets of folding solar panels that were manufactured incorrectly. But the upshot being is that they generate a significantly higher open circuit voltage so I put diodes on the 36 cell modules to stop any chance of a backfeed here. This is mainly as a protection when this array is isolated from the system, as when it's feeding in backfeeds would be non existent or negligible. One of these days I will separate the 48 cell modules and put them on their own buck converter or MPPT charge controller to get a little more out of 'em. Then the diodes can be rid of. :)
Adam I have 96 volt, 2 panels in series, 8 in parallel, will the Ideal Diode Module work at this voltage.I have a 24v (28.8v) LifePO4 batteries 120Ahr. Before I order anything.Thanks Peter.
Yeah - I may need to investigate this further. Assuming half the energy is lost (one down out of two) and we’ve got the voltage drop of two or three bypass diodes, I can’t explain why it was quite that low.
Adam Welch In series connection - did you have a bypass diode? If not, the current had to pass through the 30 -some series connected elements in the shaded panel. That means added drop in the effective voltage and the power available was less than half. You might check the individual panel voltages when they are in series and one is shaded. Or am I barking under a wrong tree?
Off topic. Solar related. If you get a chance, in a video would you explain how an MPPT charge controller controls a solar panel’s voltage to get it to its maximum power point.
vgamesx1, Thanks. I see they have MOSFETs and DC to DC converter and huge Inductor, and the idea that it needs to vary the load to get the needed voltage from the panel. I was hoping for a bit more description on how the parts interact to do that.
voltage X amps = watts. Mppt changes the load (current draw) using an adjustable buck converter to achieve the highest voltage X highest current it can for maximum watts. If it draws too much current from the panel, the panel voltage sags, thus watts can go down.
ratgreen, Adjustable buck converter! So that is what varies the load! Interesting. I know almost nothing about them - just what a buck converter does. My area was super computer design - bit twiddler;-). Understand digital - don’t know jack about power. Thanks for the seed!
@@mikemead2315 Well I'm afraid I can't help you much there but the cheapest/easiest way of doing MPPT is to monitor the input voltage and reduce charge current if it drops, have a look at Adam's CN3722 video here: th-cam.com/video/liYZ5pYOZDE/w-d-xo.html Also have a look at the bq24650, does the same thing but from TI.
I'm not that impressed with MPPT. It is not a thing that actually exists. There are many ways to determine power point. Often software programs can get lost as there is a mini power point every 10V in severe shading. I have a 60V string and sometimes I see the controller drop to 17V and stay there. Disconnecting the panels for a moment resets the system to normal voltage and higher power. This commonly happens when there is a sudden change in load. I have a number of systems that run MPPC constant voltage. That NEVER gets fooled.
*I doubt even the "ideal diodes" are good idea.* Did you measure the losses in the ideal diode? Connect the panels in parallel one with the diode in series, one without and check the power delivered by each. I don't know exactly what MOSFET they are using, but it will very likely be a P-channel-MOSFETs which have worse on-resistances than the N-channel ones. With typical values (a few tens of mOhm) you will lose around 100mW at the 2A that the panels are delivering. So 200mW in total for both panels. This is a loss that is always there. Compare that to the 400mW of backfeeding. That is a loss that is only there when one panel is completely shaded. So unless you often have the case that one panel is shaded while the other isn't (I would question the location of the setup in that case though), there does not seem to be much point in installing a diode, ideal or not.
I said I think the best solution - where possible - is to reduce the shading so I agree with your positioning statement. This was an extreme test. I agree too that despite the name, there isn’t an ideal solution. However I’m not in the shed to be able to test this, but... According to the datasheet the mosfet has a typical on resistance of 16mOhms. So the voltage drop across the module at 2 amps should be about 0.032 volts - but then we have to remember that there are two mosfets in parallel and therefore the resistance is halved. So theory has it that’s about 0.016 volts. I make that 32mW of loss.
Its an interesting topic Adam. There is another solution depending on your needs and the degree of shading that you suffer. My system has 2 banks of 2 panels (2 panels in series to lower transmission losses). The 2 banks are wired in parallel (with blocking diodes). Ideally I would have had a separate controller for each bank - indeed, I've seen an installation with a charge controller per panel and the current trend is towards microinverters. I've had my installation for 10 years now and now that the price point of charge controllers has reduced (along with _some_ having appropriate controls and profiles for LiFePo4) it may be time to reconfigure. That said, I'm still pretty happy seeing 40+ amps charging the batteries after dealing with the fridge, freezer and other loads.
I had a robust argument with a few members of a forum I belong to re. panels in parallel. I advocated fitting a Schottky diode in series with each panel to prevent reverse feeding in shade. The others said don't bother. I conceded defeat and left the diodes out, after a lot of head-scratching.
My logical conclusion goes like this:
1. None of my panels will ever be covered by a totally opaque sheet of anything.
2. The graph on the manufacturers data sheet shows that the voltage output of the panel is almost constant from max. insolation to min. insolation, only the current varies.
3. In almost any light, with no load, the voltage is almost the same, whether unshaded or translucently occluded.
4, Once the panels are illuminated, the MPPT regulates the terminal voltage to what it finds best, always below Voc
5, Therefore, if one panel is fully lit, and the others are cloud/tree shaded, the voltage will stay the same across each panel, but the current each panel contributes to the total will be in proportion the the illumination.
6. With one of my panels behind a tree shadow, and the rest fully illuminated, the Voc of the shadowed panel is higher than the Vmp of the other 5.
7. Connecting the shadowed panel in parallel with the others it's voltage drops to match, and marginally increases the total current by a few mA.
My experiments proved my original argument on the forum wrong, diodes unnecessary unless the shadow is a cardboard sheet on the surface of the panel. I ate humble pie; the money saved was worth a pint, and 25W loss on a cool bright day.
Just my tuppence worth!
Solar panels are just a bunch of crappy diodes in series. Much better than they used to be. Conduction should not happen until voltage gets significantly above the open circuit voltage. In fact, if you have snow on your panels that can be melted off by back feeding voltage as long as the current is less than the panels rated voltage.
Just purchased 5 ideal solar modules for my system, thanks for this idea.
Most solar panels have a 1000V diode 1007 kind which makes 0.7V drop or more. Obviously a 40V Schottky Diode can do better given not too many are connected in series. I think the idea of using a 1000V one is also to make this more robust against voltage spikes as they might occur in lightning strikes (which won’t hit your house but something else close enough to have an impact). The mosfets chosen are probably also low voltage kinds with similar risks in case of a spike or so. In case of dozens of solar panels on the roof who wants to periodically check diodes and diagnose might be difficult. The waterproof box probably isn’t waterproof after opening it up twice...
Guess what, Schottky diodes leak too. Especially when hot. Lower voltage rated ones have lower forward voltage than ones with a higher PIV. I was hit by lightning and it took out a number of bypass diodes protecting the panel. Thank heavens for cheap diodes. They were easily replaced and the panels are still working.
I agree, maybe a quick reaction and high voltage solution makes sense for safety and lower maintenance
Hey, great video again. To my understanding panels themselves are bypass-protected so that a shaded panel does not have to pass the current through it's cells during shading of a panel in a series string. This mimics removing a panel from the series and avoiding the shaded panels resistive properties.
However, when "overloading" a solar chargers input with more than one string in parallel, at most with one each facing east, south and west, there'll be a situation where entire strings of panels are shaded. Especially when the voltage of the strings, or difference of voltages of single parallel panels are great enough, damage may occur, to my understanding again. I'm about to combine a 21Voc and a 41Voc panel in parallel on a single MPPT and will be using blocking diodes.
To demonstrate, try connecting a 230V lightbulb to a 12Vdc source and observe current; nothing much happens... Until the voltage becomes high enough to overcome the resistance and the wire starts to glow. I'd avoid making panels into lightbulbs...
Panels do not contain blocking rectifying on top of bypass, because it would lower the ratings of the panel. They also want to avoid adding cost because they want to be the cheapest on the market... But when you go farther away from the sun to say Finland, you either have to move panels around according to time of day or get more of them in different directions in order to maximize output and usage of an expensive inverter in order to make things as affordable as possible.
Maybe there's a sequel to this topic? Thanks and keep up the great work!
never heard of an ideal diode module. Pretty cool.
I built my own. It's just three transistors,three resistors, and two diodes.
Couple of points: most panels seem to be one long series string of 0.5v cells and shading just one cell can seriously reduce output. If one cell is a bit dirty, what happens to output ? Secondly when doing this sort of experiment it would be better to find a constant and thus repeatable load rather than just a battery on charge.
Within panels the same thing happens just on a smaller scale. So yes, one "dirty" cell will reduce output more than just having one cell less.
The load is fine in this setup as long as the batteries aren't the limiting factor. MPPT tries to get the most power out of the cells in any case. And comparing measurements more than a few minutes apart is pointless anyway, because the sun/haze will have changed more and uncontrollable than anything else in the setup.
So this is not a video on how to shade panels! Ha-ha great video, alot of ppl think a little shadow on a panel is nothing!! In fact it is a big nothing when it comes to power!!
JAMES! Since the 70's not necessarily.
The parallel connection is very interesting. In this case the consumption of the shaded panel is tiny but when you have strings of 23 modules of 300w each in parallel it may be required to use string fuses to avoid destroying the strings when heavy shading occurs as the Voc will be close to 1kV and the current capabilities are 9-10A... Each manufacturer will give the maximum series fuse rating in amps which is usually 15A or 20A. Usually string fuses are required when more than 2 strings are installed in one MPPT in parallel and they are there to protect the string from being destroyed. Also, string diodes are not used anymore.
You can test for night time back feeding, with a IR , night vision camera, they will have a slight glow, in IR,
I did not know that - I might try to get my hands on a camera to try that. Cheers
So on my pair of panels on my shed roof, one often gets a bit of shade from a neighbour's tree, and later in the day a shadow from next door creeps across them. Looks like I need to parallel them up - even without diodes/ideal diodes, the loss was much less on parallel than serial!
I put mine in parallel as the house shades one of my two large panels from about 3pm, which when in series was causing a considerable loss.
@@ganey I've just rewired my pair to parallel. Bit of a poor day weather wise for solar panels though so will have to wait for the next sunny day to really compare!
It's hard to compare with the randomness of our weather but my pair of panels are definitely working better since changing them to parallel. They produce power later into the day and the peak power is around 10% higher than I've ever achieved before!
@@stevec00ps I'm glad it's working better for you in parallel!
Hi Adam and all
Thank you for the demonstration of the ideal diode,
One question that haven’t been mentioned, is the effect of putting miss matched panels together.
And how to avoid the same problem, for example putting a 40volt panel in parallel with a 20volt panel. in this instants the problem you discussed would be much worst, and best avoided by trying to match panels (in parallel for VOLTAGE, and in series for CURRENT) would you agree ?
This is something I’m trying to tackle with expanding my solar array and deciding whether or not to add Schottky diode in one of my two strings.
Many thanks for the video.
Very helpful video 👍
How much power could be gained, by ditching the diodes, for the ideal diodes? On a system with about 10, 300w panels? That's about 5volts, total!!
I guess it could help a small system, great video...
hy brother i want to ask about diode ideal..can use it to mppt esmart..i see spesification mppt dont connect solar panel to mppt without battery connected
Seems like in series its still better becauee it produces more while not shaded but about the same as in parallel (25-30W) when shaded.
A PMOS, a 100 Ohm gate resistor, and an optional 18V Zener to protect the gate against over voltage (20V), will do the same job.
I think the 6-pin IC that is controlling the 2 PMOSFETs, is a cheap Chinese clone version of something like an LTC4359, that has been repackaged in a
6-pin package.
The real ones are 8 pins, but 3 pins have no internal connection. I think the cheapest, with least resistance, is a fuse and diode combo.
If the power is reversed, the diode across the power will conduct and blow the main fuse.
Thanks for this. That looks like an experiment I could look at doing :-)
Have had this video on my must watch list for a while, glad I did. Very interesting as usual. But, what about low light? Realistically most solar setups are capable of producing "more than enough power in ideal conditions so the differences in setup are not really important until winter when the sun is low and clouded. Would love to see your opinions on low light solar generation variations
dont all panels have internal diodes....And are they not a high resistance so not really a burden to other parallel panels?
I need to investigate this as i have not seen this parallel problem before/in my setup.
edit ok you just did what i was going to do, and 1/3w loss, meh, i'll live with that
Off topic. Solar related. Eco Worthy now also has a 40 amp MPPT charge controller that seems to have bluetooth communications ability with a SmartPhone. Years ago you reviewed and used their 20 amp MPPT, and I own the 20 Amp version. I have moved on, but I thought their 20 amp MPPT was a great value. I was wondering if their 40 amp MPPT was competitive in the value market.
Voltage drop on the ideal module?
For a tipical MOSFET (NTB5860NL), the On resistance is 0.003 Ohm so at 1.9 A it will be *6 mV*
This is because when the transistor is on it behaves like a resistor, the parameter is Rds(on) if you look on the datasheet.
hey thanks, would you be able to duplicate the test with partial shading? I'm looking to set up a test of single axis versus dual axis small array...
Thanks!!!
Thanks Adam. I've only ever made small solar panels from individual cells, and always fitted a suitable Germanium diode (back in the 1970's and 80's) in series with each panel, to avoid reverse-bias situations. Do commercial panels not have similar protection? If not, then it's sensible to fit your own diodes, or MOSFETs, in a parallel array, as you describe here.
I’ve not seen a panel supplied with an inline diode. Usually this is left to the solar charge controller which will have protection on the solar input, but of cause this doesn’t help when paralleling up panels which get shaded. Most have bypass diodes for when they are setup in series though.
Each of my panels is connected to its own microinverter so, in theory, I should benefit from taken the diodes out of the panel altogether, is that right? Also, I'm considering making a solar tracker as I only have 6 250w panels and I believe it is cheaper to make more power from the panels I got rather than buying more panels. Have you considered making a solar tracker yourself?
Those in the panel are bypass diodes. They prevent damaging a cell that is shaded by reverse voltage. If you don't have severe shading they do nothing either way. Remove tham and you will destroy the panel. If you have the room, panels are way cheaper than tracking these days if not buying battery voltage panels.
Very informative video Adam, thanks...
Excellent video.... 👍😎🖖
How to find the Mosfet module you are using in this video? I want to use my panels.
Links in the description - it’s an ‘ideal diode’
Thanks for this interesting video. However, I'm confused or maybe missing something obvious.
With two panels in series you measure 75W, but when in parallel only get out 72% (2x27W=54W) power in full sunlight (or was the measurement taken in different solar conditions).
Why such a difference? Is serial solar panels setup always more power efficient?
That's because it's not a lab test where you can keep the "sun" constant. The position of the sun changed, maybe there was more haze, that already makes a big difference. Comparing measurements a minute apart might be fine, anything more than that not really.
Yeah there was some time between these tests. Conditions were warm and sunny but there were some clouds around. It’s hard to get the exact same conditions for each and every test, but actually it wasn’t really needed for this one - it was more about comparing shaded and unshaded panels.
just opened one of my 38v 300w panels, to find its 2 series 19v 150w panels..... the 'first' panel has a diode ''over it'' to prevent reverse voltage, and the 'second' panel also has this, and the + of the first is connected to a diode which goes to the - of the second... odd but functional way. there only cheap silicon diodes, makes me wonder why decent manufacturers dont change them for something better but i suppose its how its been for 30 years or more
Interesting, thank you.
I don't think it's worth the trouble to put ideal diodes on every panel in parallel, and certainly not conventional diodes as they will end up wasting more power than they save. In my setup that would require a *hell* of a lot of ideal diode modules! However I am considering building my own high powered versions to replace the main reverse blocking diodes in the system. One day I'll get a Round Tuit.
On the other hand, parallel panel configurations of solar modules are superior if they are in a situation where any part of the array gets shaded during the day. And of course, the only way if you want to have different directional facing modules on the same controller. Five weeks past the Winter Solstice here - the Sun is coming back nicely. :)
Hey Xan. Glad you’ve got through another winter and the sun is returning. I agree, in most cases the diodes are unnecessary and reducing or removing the shading is always going to be the best option. Plus I doubt anyone would experience quite the level of shading I tried here! I tried to cover this in the conclusion at the end of the vid. Cheers for your input as ever.
@@AdamWelchUK Yes, in certain situations diodes are useful. I actually do have some, one of my arrays has a combination of standard 36 cell 12 volt nominal modules and also three bizarre 48 cell modules. These are not grid tie panels, they are actually sets of folding solar panels that were manufactured incorrectly.
But the upshot being is that they generate a significantly higher open circuit voltage so I put diodes on the 36 cell modules to stop any chance of a backfeed here. This is mainly as a protection when this array is isolated from the system, as when it's feeding in backfeeds would be non existent or negligible.
One of these days I will separate the 48 cell modules and put them on their own buck converter or MPPT charge controller to get a little more out of 'em. Then the diodes can be rid of. :)
Good info, thanks
If you are so concerned about shading you need these solar panels:
th-cam.com/video/fIKy7X3Jnlk/w-d-xo.html
Adam I have 96 volt, 2 panels in series, 8 in parallel, will the Ideal Diode Module work at this voltage.I have a 24v (28.8v) LifePO4 batteries 120Ahr. Before I order anything.Thanks Peter.
The module I used can be used with panels up to 28 volts. I guess other modules might go higher.
@@AdamWelchUK Thanks Adam. Peter
Why was there such a massive loss in series. Output dropped to a third when 50% was covered.
Yeah - I may need to investigate this further. Assuming half the energy is lost (one down out of two) and we’ve got the voltage drop of two or three bypass diodes, I can’t explain why it was quite that low.
Adam Welch
In series connection - did you have a bypass diode? If not, the current had to pass through the 30 -some series connected elements in the shaded panel. That means added drop in the effective voltage and the power available was less than half. You might check the individual panel voltages when they are in series and one is shaded. Or am I barking under a wrong tree?
What is that device in the middle at 9:25 ?
It’s the MT50 remote meter for my EPEver Tracer A solar charge controller. There’s a link in the description :-)
Woo 😁 cheers.
Off topic. Solar related. If you get a chance, in a video would you explain how an MPPT charge controller controls a solar panel’s voltage to get it to its maximum power point.
vgamesx1, Thanks. I see they have MOSFETs and DC to DC converter and huge Inductor, and the idea that it needs to vary the load to get the needed voltage from the panel. I was hoping for a bit more description on how the parts interact to do that.
voltage X amps = watts. Mppt changes the load (current draw) using an adjustable buck converter to achieve the highest voltage X highest current it can for maximum watts. If it draws too much current from the panel, the panel voltage sags, thus watts can go down.
ratgreen, Adjustable buck converter! So that is what varies the load! Interesting. I know almost nothing about them - just what a buck converter does. My area was super computer design - bit twiddler;-). Understand digital - don’t know jack about power. Thanks for the seed!
@@mikemead2315 Well I'm afraid I can't help you much there but the cheapest/easiest way of doing MPPT is to monitor the input voltage and reduce charge current if it drops, have a look at Adam's CN3722 video here: th-cam.com/video/liYZ5pYOZDE/w-d-xo.html
Also have a look at the bq24650, does the same thing but from TI.
I'm not that impressed with MPPT. It is not a thing that actually exists. There are many ways to determine power point. Often software programs can get lost as there is a mini power point every 10V in severe shading. I have a 60V string and sometimes I see the controller drop to 17V and stay there. Disconnecting the panels for a moment resets the system to normal voltage and higher power. This commonly happens when there is a sudden change in load. I have a number of systems that run MPPC constant voltage. That NEVER gets fooled.
Bypass diodes should be in the opposite direction :-)
*I doubt even the "ideal diodes" are good idea.* Did you measure the losses in the ideal diode? Connect the panels in parallel one with the diode in series, one without and check the power delivered by each. I don't know exactly what MOSFET they are using, but it will very likely be a P-channel-MOSFETs which have worse on-resistances than the N-channel ones. With typical values (a few tens of mOhm) you will lose around 100mW at the 2A that the panels are delivering. So 200mW in total for both panels. This is a loss that is always there. Compare that to the 400mW of backfeeding. That is a loss that is only there when one panel is completely shaded. So unless you often have the case that one panel is shaded while the other isn't (I would question the location of the setup in that case though), there does not seem to be much point in installing a diode, ideal or not.
I said I think the best solution - where possible - is to reduce the shading so I agree with your positioning statement. This was an extreme test. I agree too that despite the name, there isn’t an ideal solution. However I’m not in the shed to be able to test this, but... According to the datasheet the mosfet has a typical on resistance of 16mOhms. So the voltage drop across the module at 2 amps should be about 0.032 volts - but then we have to remember that there are two mosfets in parallel and therefore the resistance is halved. So theory has it that’s about 0.016 volts. I make that 32mW of loss.
YUP, and now you have IR loss problems with your perfect diodes.
The diode nsme is Schottkey not Schockey. ;)
Great info
Thanks Colin